ICES J. mar. Sci., 49: 23-44.1992 The meiobenthos of the North Sea: density, biomass trends and distribution of copepod communities* R. Huys, P. M. J. Herman, C. H. R. Heip, and K. Soetaert Huys, R., Herman, P. M. J., Heip, C. H. R., and Soetaert, K. 1992. The meiobenthos of the North Sea: density, biomass trends and distribution of copepod communities. - ICES J. mar. Sci., 49: 23-44. During a synoptic survey carried out in April-May 1986, 171 localities were sampled in the North Sea as delimited by the Straits of Dover in the south and approximately by the 100m isobath in the north. Meiobenthos included Nematoda, Copepoda, Turbellaria, Gastrotricha, Polychaeta, Oligochaeta, Priapulida, Kinorhyncha, Ostracoda, Halacarida, Isopoda, Tanaidacea, Bryozoa, Cnidaria, Sipunculida, Echiurida, Nemertini and Tardigrada. Nematodes were the dominant group in virtually all stations, their densities ranging from 61 to 4167 individuals. 10em:". Only in the Southern Bight, where nematode numbers were low, did harpacticoids sometimes represent the dominant meiobenthic taxon. There was a tendency for nematode (and total meiobenthos) density to increase towards the north. A total of278 copepod species belonging to 105genera and 22 families were identified. Over 40% of the species were new to science; new taxa were found particularly among the interstitial families which were most important in terms of species diversity. Copepod density decreased rapidly to the north and this trend was followed by diversity. Individual ash-free dry weight (AFDW) was determined for 98 species of copepod. Total biomass reached a peak in the south (low mean individual AFDW, high density) and in the north (high mean individual AFDW, low density), but was low in the Central North Sea where the copepod communities were impoverished both qualitatively and quantitatively. Using the classification technique TWINS PAN (two-way indicator species analysis), it was impossible to define meaningful clusters (TWIN groups) on the basis of the 18 major meiobenthic taxa. However, seven distinct communities could be recognized on the basis of the copepod composition: (1) TWIN A largely coincided with the Southern Bight and showed high densities of predominantly interstitial species (Cylindropsyllidae, Paramesochridae, Cyclopinidae) and a few characteristic taxa from coarse sediments; (2) TWIN B was found in the coastal zone of the Netherlands, Germany and Denmark, and in the Dogger Bank, and was dominated by large Ectinosomatidae and Ameiridae, and by interstitial Leptastacidae; (3) TWIN C represented an impoverished community north of the Dogger Bank and consisted of large mud-dwelling species belonging to the Diosaccidae, Laophontidae and Arneiridae; (4) between the Scottish coast and Norwegian Deeps and in the Silver Pits Zosimidae, Cletodidae and Idyanthidae were the most important families (TWIN D); (5) TWIN E grouped the Norwegian Deeps, Devil's Hole and Farne Deep and showed a typical deepwater fauna represented by Ancorabolidae, Cerviniidae, Stenocopiinae and bathyal c1etodid genera. Two minor clusters (a, p) coincided with the Dutch Wadden Sea (I station) and the river outlets (Thames, Wash, Meuse/Scheldt) whose meiobenthos is subject to pollutants. Canonical Correspondence Analysis (CCA) clearly separated the five major twin groups. TWIN A-C were significantly correlated with sediment and could be arranged along a gradient of decreasing median grain size and increasing silt/ clay content. TWIN D was clearly related to latitude whilst TWIN E showed a clear preference for depth. Key words: North Sea; meiobenthos; copepod communities; biomass; diversity; latitudinal trends; synoptic survey. Received 22 May 1991;accepted II November 1991. R. Huys, P. M. J. Herman, C. H. R. Heip, and K. Soetaert: Netherlands Institute of Ecology, Centre for Estuarine and Coastal Ecology, Vierstraat 28, 4401 EA Yerseke, The Netherlands. *Contribution no. 500 of the Delta Institute for Hydrobiological Research, Yerseke. 1054-3139/92/010023 + 22 $03.00/0 © 1992 International Council for the Exploration of the Sea 24 R. Huyset al. Introduction Since Smidt's (1951) early work on the Danish Wadden Sea and Mcintyre's (1964) study of the Fladen Ground meiobenthos, much information has been gained on the species composition, density and biomass ofmeiobenthic assemblages. Heip et al. (1990) summarized this knowledge, resulting from 40 years of meiobenthos research in the North Sea, and concluded that only the coastal areas of Belgium, the Netherlands and Germany were relatively well known (Fig. 1). Large areas of the North Sea have not been investigated and in particular the lack of basic data from deeper localities makes it impossible to present a comprehensive picture of North Sea meiobenthos. Even for the easily accessible, shallow areas of the Central North Sea such as the Dogger Bank and the eastern sandbanks (Jutland, Little Fisher, Turbot), information on meiobenthos remains fragmentary. There were several attempts to define biological regions within the North Sea but these were all based on planktonic communities or more recently on macro benthic infauna and epifauna (Basford et al., 1989, 1990; Eleftheriou & Basford, 1989). The question remains whether meiobenthos or its component major taxa can also be applied to define communities on a large geographical scale. A synoptic survey of the North Sea benthos was organized during April-May 1986 in the area delimited by the Straits of Dover in the south and approximately by the 100 m isobath in the north. This sampling programme involved the cooperation of 10 laboratories from France, Belgium, The Netherlands, Germany and the UK. During this survey we sampled a grid of 171 localities for meiobenthos and for various physicochemical sediment parameters. This paper presents an analysis of the distribution of the major meiobenthic taxa, and assesses the major environmental factors affecting the various copepod communities. Materials and methods The sampling area The present investigation was conducted as part of the North Sea Benthos Survey (NSBS) executed in AprilMay 1986. This synoptic survey covered a total of 197 localities arranged in the ICES grid from the Straits of Figure I. Map showing bathymetry of southern and central North Sea. Meiobenthos of the North Sea 25 Figure 2. North Sea map showing sampling localities of North Sea Benthos Survey. Filled circles indicate localities that were not sampled. Dover in the south to approximately the 100m isobath in the north and from 2°30' W to 8° IS' E. Only 171 stations were sampled for meiobenthos (Fig. 2), and a complete list of the sampling data was reported elsewhere (Anonymous, 1986) and is available from the ICES database on request. Stations not being sampled included localities along the Danish coast (102,144,164,165,174,176,183) and in the German Bight (64, 73, 82), in the Dutch Wadden Sea (26), in the mouth of the Humber and The Wash (27, 28, 32,47), and in the Fladen Ground off the northeastern coast of Scotland (188-193). Some are shallow localities and were for that reason not accessible by the research vessels used. Others were not sampled because of the unfavourable weather conditions at the time of sampling. Bottom samples were taken with a Van Veen grab (0.184 m") or preferably with a box corer (0.068 m2) or with both sampling gears. Usually four subsamples (10 em") were taken with perspex cores, but only one subsample was provided for the localities sampled by the Nederlands Instituut voor Onderzoek der Zee (NIOZ, Texel) and by the Forschungsinstitut Senckenberg (Wilhelmshaven). Some of the latter stations were either sampled by both or by other laboratories as well so that for 134 out of 171 localities two or more subsamples (Iu crrr') were provided. Use of the Van Veen grab in sandy bottoms might cause a considerable loss of interstitial water when lifting on board, resulting in an overestimation of the relative abundance of the surface- 26 R. Huysetal. dwelling species such as copepods and an underestimation of the relative abundance of the interstitial small-sized types which often represent the bulk of the copepod community in coarse sediments. Conversely, in fine sediments where surface-dwelling species are dominant the bowwave of the Van Veen grab may push epibenthic species away before sampling. Hence, box corer samples were preferred whenever they were available. The radius of outliers is equal to the smallest or largest radius indicated below the box and whisker plot. For intermediate values the radius is interpolated between these extremes. This interpolation is linear on a log (density) scale. Concentric circles indicate data derived from two replicates taken by different laboratories. Measures of species diversity Processing The following Hill's diversity numbers (Hill, 1973) were used as measures of diversity of the copepod community and calculated on the raw data: Meiobenthos samples were fixed and preserved in a hot (about 70°C), 4% formaldehyde solution. Animals were No = the number of species in the sample (species richsubsequently extracted by centrifugation-flotation in ness); equals in this study the "species density" LUDOX TM® colloidal silica (Heip et aI., 1985) or by (number of species per unit area = 10cm') as all decantation using a 38 11msieve when the sediment proved copepods in the sample were identified. too coarse. Copepods were first picked out of the samples, N =exp (H'), with H' the Shannon-Wiener diversity 1 since their identification often required dissection. The index calculated with natural logarithms. remaining meiobenthos was counted and identified at high taxonomic level (phylum or class) under a stereo- Classification scopic microscope, after staining with Rose Bengal. All Two-way indicator species analysis (TWINSPAN) (Hill, copepods were identified down to species level. 1979)was used to classify the stations for which copepods Individual dry weights of copepods were determined on were examined. TWINSPAN is a polythetic divisive techa Mettler M3 electronic microbalance (± Oi l ug). Batches nique that first performs a primary ordination on the of 20-150 (according to size) specimens belonging to the samples by reciprocal averaging, and then uses this ordisame species were rinsed three times in double-distilled nation to obtain a classification of the species according to water, dried for 2 h at IlDoC, cooled in a desiccator, and their ecological preferences. The synecological relations weighed. Only adults were selected for weighing and no of the various species can be expressed in an ordered twoattempt was made to separate the sexes because of the way table constructed from the sites-by-species matrix. paucity of some of the species. Females and males were TWINS PAN also identifies one to several differential evenly represented in every batch. However, females species (=indicator species) which are particularly diagcarrying eggs were not selected since it was found that the nostic of each division (twin group) in the dendrogram presence of an ovisac could increase ash-free dry weight (indicator ordination). In addition, preferential species (AFDW) by over40%. which appear to prevail in one side of a dichotomy may be selected as well. In order to handle quantitative data as well, each species abundance is replaced by the presence Mathematical techniques - Statistical analysis of one or more pseudo-species. Each pseudo-species is Distribution maps defined by the minimum abundance of the corresponding Maps were produced using the program DIHOMAP species (cut level); thus the more abundant a species, the (Herman & Braat, 1991).This program addresses a data- more pseudo-species will be defined. base and maps the densities of species and higher taxoOrdination nomical groups. The radius of the symbols is proportional to the log-transformed density. Details of this are given by In many cases it is useful to parallel TWINSPAN classification with an ordination. The ordination technique used the associated box and whisker plots. The box is drawn between the first and third quartile of is canonical correspondence analysis (ter Braak, 1986). the (log-transformed) data. The position of the median on The resulting ordination diagram not only expresses the the log-transformed scale is indicated by the vertical bar. pattern of variation in species composition, but also reflects the major relations between the species/stations The back-transformed value of the median is indicated above this line. The whiskers are the lines extending from and each of the given environmental variables. The both sides of the box to the left and right. They extend to stations of each twin group are positioned as points (symbols) in the CCA diagram. The environmental variables the most extreme observation lying within the boundary quartile value ± 1.5 times the interquartile distance. Their are represented by arrows and can be interpreted in conback-transformed value is indicated below the box and junction with the station points. Each arrow determines whisker plot. Observations falling beyond the whiskers an axis in the biplot and the station points must be proare called outliers. Their number (not their value!) is given jected onto this axis. The order of the projection points corresponds approximately to the ranking of the weighted by the numbers to the right and left of the whiskers. Meiobenthos (a) of the North Sea 27 ( b) Frequency: III Frequency: 202 co 22 3 4 9 0 0 0 0 76 88 • • • • •• •• • • • • •• • •• • • ..-..-------~ Figure 3. Log of total density of the major soft-bodied taxonomic groups of the meiobenthos in the North Sea. (a) Turbellaria. (b) Gastrotricha. Between the minimum and maximum values indicated under the box and whisker plot, the radius of the symbols is proportional to the log-transformed density (see text for details). averages of the species with respect to a particular environmental variable. Environmental variables with long arrows are more strongly correlated with the ordination axes than those with short arrows, and therefore more closely related to the pattern of variation of species composition shown in the ordination diagram. Arrows that point in the same direction indicate positively correlated variables, perpendicular arrows indicate lack of correlation and arrows pointing in the opposite direction indicate negatively correlated variables. Results Total meiobenthos Meiobenthos included Nematoda, Copepoda, Platyhelminthes (Turbellaria), Gastrotricha, Polychaeta, Oligochaeta, Priapulida, Kinorhyncha, Ostracoda, Halacarida, Isopoda, Tanaidacea, Bryozoa, Cnidaria (Hydrozoa), Sipunculida, Echiurida, Nemertini and Tardigrada. Nematodes were nearly always the dominant group in the meiobenthos, their densities ranging from 61 to 4l67ind. lu cm"? (x=759ind. lu cm"). Only in the Southern Bight were harpacticoids sometimes as abundant as nematodes or were even the dominant taxon of the meiobenthos. In the remaining localities nematodes accounted for at least 85% of the meiobenthos; Harpacticoida or Turbellaria (and in a few cases Gastrotricha) ranked second in abundance. The other groups were present especially in medium coarse or coarse sands but they were far less common than the main taxa. Figure 3a and 3b illustrate the distribution and density of the two dominant soft-bodied meiobenthic taxa, the Turbellaria and the Gastrotricha. The central part of the Southern Bight (south of 53°30' N) contained stations with high numbers of 28 R. Huyset interstitial copepods and relatively low nematode densities. This community extended to the coastal zone of Belgium and The Netherlands but was apparently absent in the shallow offshore area of Britain although the sediment type was virtually the same (median grain size averaged 250-300 urn). The Southern Bight community seems to be unique for the entire North Sea; values of the nematode:copepod ratio were low and ranged between 16 and 25. Only a few localities along the western coast of Denmark and around the Isle of Sylt gave indication of a similar nematode:copepod ratio (Fig. 4a). There was no monotonic trend discernible for the N/C ratio with latitude (Fig. 4b). Although coarse sediments generally favour the development of a characteristic interstitial fauna it is striking that the mesopsammic harpacticoids were not typical for the German Bight and the west coast of Denmark. Particularly in the entrance to the Skagerrak where very coarse sediments were found, these tiny copepods seemed to be outnumbered by other ecotypes, i.e. epi/endobenthic copepods. Here, high densities of gastrotrichs were recorded. The distribution of the kinorhynch genera was clearly related to the median grain size. Four genera were recorded, and these can be assigned to two ecological groups. Echinoderes and Semnoderes are typical representatives of sandy sediments and particularly the latter genus is known to inhabit coarse substrata. These genera were found only in the Southern Bight and in the entrance to the Firth of Forth (Fig. 5). The latter record is not surprising as the sediments in this isolated area were fine or medium sands in contrast to the very fine sediments (below 200 urn) of the surrounding waters. A similar case was found in the German Wadden Sea (station 46). Conversely, species of the genera Pycnophyes and Kinorhynchus were recorded in the central part of the North Sea and never in the Southern Bight. Their distribution was largely confined to the eastern part of the central area (Fig. 5). It is known from the literature that these genera are associated with very fine sediments. The same area was also characterized by the occurrence of Priapulida (larvae of Priapulus caudatus), a taxon that was entirely absent in the Southern Bight and occurred only sporadically in the western part of the North Sea. Nematodes, generally the most abundant meiobenthic taxon in marine sediments, became even more abundant with latitude (and thus depth) up to 53°30' N. From this latitude was a tendency for nematode density to decrease towards the north, but the trend was not linear (Fig. 6a). The relation between total meiobenthos density and latitude fitted the picture obtained for the nematodes (Fig.6b). TWINSPAN was applied to classify the stations on the basis of absence or presence of the 18 major meiobenthic taxa but did not produce any meaningful groupings. One of the shortcomings of TWINS PAN is that taxa that are al. (0) Frequency: 194 8-rn58 1701 • 52 53 54 55 56 57 58 59 Latitude Figure 4. (a) Log of nematode:copepod ratio. Between the minimum and maximum values indicated under the box and whisker plot, the radius of the symbols is proportional to the log-transformed NjC ratio (see text for details). (b) Trend of nematode:copepod ratio with latitude. Per degree latitude nematode:copepod ratio of all stations falling in that zone is averaged. Error bars indicate standard errors of the mean. rare and moreover occur in stations with low diversity determine the arrangement of the stations. For example, 29 Meiobenthos of the North Sea Frequency: 14 42 (a) I 12 o----~ 3 10 8 4 6 0 4 ro Q ~ 2 51 OJ 'E 15 z 14 o o 52 ~---L_ 53 54 59 (b) 13 12 II 10 9 8 7 6 5 51 Latitude Figure 6. Trend of total density with latitude. Per degree latitude the density of all stations falling in that zone is averaged. Error bars indicate standard errors of the mean. (a) Nematoda. (b) Total meiobenthos. ~--------~ Figure 5. Log of total density of the Kinorhyncha in the North Sea. Between the minimum and maximum values indicated under the box and whisker plot, the radius of the symbols is proportional to the log-transformed density (see text for details). Empty circles denote mud-dwelling genera (Pycnophyes, Kinorhynchus), filled circles denote sand-inhibiting genera (Semnoderes, Echinoderes). the first TWINSPAN dichotomy was primarily determined by the Cnidaria and the Tardigrada, the second division basically by the minor taxa Isopoda and Halacarida, etc. Of the major groups only the various pseudo-species of the Gastrotricha seemed to contribute in shaping the early subdivisions of the dendrogram. Copepoda The fauna The harpacticoid community predominantly included adults and post-metamorphosed stages (copepodites). Examination of the copepod fauna from one sample (10 ern -2) of each of the 171 selected stations resulted in a total number of 7710 individuals. All copepods were identified to species level. The complete faunal data are available from the authors on request. A total number of 278 species belonging to 105 genera and 22 families were identified. Surprisingly, 121 species (43.5%) turned out to be new to science. A high number of novel species was recorded for the Paramesochridae (27), Cylindropsyllidae (24), Ectinosomatidae (17) and Ameiridae (14). For both the Paramesochridae and the Cylindropsyllidae, the North Sea Benthos Survey resulted in a doubling of the species number for the North Sea area as delimited by the ICES boundaries. The examination of the interstitial copepod fauna revealed also nine new genera. The vast majority of the fauna belonged to the Harpacticoida. The Cyclopoida were represented by the primarily mesopsammic family Cyclopinidae (six species). An additional number of calanoids and planktonic cyclopoids (Oithonidae) was omitted in the analysis. Qualitatively important families were Paramesochridae (44 species), Cylindropsyllidae (38), Ameiridae (35), Ectinosomatidae (34), Cletodidae (26), Laophontidae and Diosaccidae (21). A total number of 105 genuinely interstitial species were recorded, including the Paramesochridae, Cylindropsyllidae, Cyclopinidae and small-sized representatives of the Ameiridae, Canthocamptidae, Diosaccidae and Ectinosomatidae. The remainder consisted 30 R. Huysetal. Table I. Total number of specimens found and the number of occurrences in the total data set for the top 50 species of the survey. Family Leptastacidae Leptastacidae Cylindropsyllidae Leptastacidae Cylindropsyllidae Ectinosomatidae Paramesochridae Paramesochridae Diosaccidae Diosaccidae Diosaccidae Paramesochridae Cylindropsyllidae Cylindropsyllidae Ectinosomatidae Paramesochridae Ectinosomatidae Ameiridae Cletodidae Cyclopinidae Leptastacidae Ectinosomatidae Ectinosomatidae Idyanthidae Leptastacidae Diosaccidae Ameiridae Diosaccidae Ameiridae Paramesochridae Paramesochridae Cylindropsyllidae Paramesochridae Leptastacidae Ectinosomatidae Paramesochridae Thalestridae Paramesochridae Ectinosomatidae Cylindropsyllidae Cletodidae Longipediidae Ectinosomatidae Cylindropsyllidae Ameiridae Cylindropsyllidae Ectinosomatidae Cylindropsyllidae Cletodidae Paramesochridae Species Leptastacus laticaudatus Paraleptastacus espinulatus Evansula pygmaea Arenocaris bifida Leptopontia curvicauda A renosetella germanica Kliopsyllus holsaticus Intermedopsyllus intermedius Psammotopa phyllosetosa Protopsammotopa norvegica Typhlamphiascus con/usus Kliopsyllus constrictus Boreopontia heipi Stenocaris kliei Halectinosoma herdmani Paramesochra mielkei Pseudobradya beduina Interleptomesochra eulittoralis Stylicletodes longicaudatus Metacyclopina brevisetosa Paraleptastacus holsaticus Ectinosoma melaniceps Bradya scotti Idyanthe pusilla Leptastacus sp. 2 Bulbamphiascus imus Pseudameira crassicornis Paramphiascopsis longirostris Proameira hiddensoensis Kliopsyllus paraholsaticus Apodopsyllus listensis Syrticolaflandrieus Wellsopsyllus gigas Leptastacus sp. I Halectinosoma propinquum Paramesochra helgolandica Pseudotachidius coronatus Scottopsyllus minor Pseudobradya minor Cylindropsyllus laevis Enhydrosoma sp. I Longipedia helgolandica Pseudobradyasp. I Stenocaris minor Ameiropsis brevicornis Cylindropsyllus remanei Halectinosoma sarsi Arenopontia sp. 3 Cletodes tenuipes Kliopsyllus sp. 4 Total number of specimens Frequency 356 294 235 199 195 132 130 116 107 98 96 95 94 82 80 80 79 78 78 74 73 71 68 65 65 60 56 55 55 54 51 50 50 49 48 47 47 46 43 42 42 42 42 42 41 41 41 40 39 39 41 42 22 32 20 19 15 17 23 16 17 II 12 13 25 13 16 20 19 15 17 II 16 II 14 15 17 22 13 13 8 12 17 12 17 7 6 II 12 10 II 8 9 13 14 9 19 9 8 7 mainly of large epibenthic or burrowing harpacticoids. Biomass Euterpina acutifrons and Microsetella norvegica are likely Determination of biomass is invaluable in quantitative ecological investigations. However, most published data give only rough estimates on total meiobenthic biomass. Direct weighing is often circumvented through calculation of the total volume, either (a) by approximating it to a particular geometric shape, or (b) by a crude division to be contaminants derived from the plankton since these species are generally believed to be holoplanktonic. Total number of specimens found and the number of occurrences in the total data set are given for the 50 most abundant species in Table 1. Meiobenthos of the organisms into a number of shapes whose volumes are summed up, or (c) by using length and width measurements and a conversion factor derived from plasticine models made from scale drawings (Gee & Warwick, 1984). Obviously, application of biomass conversion factors, based on average individual weights, to material from which they were not originally calculated can result in significant errors. Reliable biomass values can be obtained if the average individual dry weight of each species is measured. In practice, this approach is far too time-consuming for most ecological work and direct dry weight measurements of the rarer species can easily be over- or underestimated when the number of specimens in the batch is insufficiently high. Literature values on individual ash-free dry weights of copepods are scarce (Goodman, 1980; van Damme et al., 1984;Herman et al., 1984;Herman & Heip, 1985)and are determined for species taken from small geographical areas where a particular aspect was under investigation. A survey of the copepod fauna on such a large scale as the North Sea offered the opportunity of compiling a checklist of specific biomass values for copepods since numerous specimens were available for many species. This list can be extended at any time when new data become available and could refine considerably the calculation of the overall copepod biomass in future investigations. The dominance of small-sized harpacticoids in the Southern North Sea made it hard to determine accurately the biomass of the copepod fraction of the meiobenthos. Hence, special emphasis was placed on the determination of individual dry weights of interstitial copepods such as the Cylindropsyllidae, Leptastacidae and Paramesochridae. A total of 98 different species belonging to 2 I families and 73 genera were weighed (Table 2). The other copepods were assigned to one of these 98 values according to their size, shape and exoskeletal properties. It has to be remarked that overall total biomass figures will usually be slight overestimates since they were based on adults only, and no attempt has been made to divide the species into different size- and biomass-classes according to their copepodid stages. These overestimates will be negligible in the Southern North Sea where mesopsammic copepods dominate. Latitudinal trends There is a distinct and significant trend for copepod density to decrease towards the north (Fig. 7a). Highest values (l81 indo 10cmr ') were recorded in the Southern Bight between 5l.so and 52°N where tiny interstitial copepods showed an overwhelming dominance in the community. Density decreased rapidly towards the Dogger Bank and reached its minimum average value (l8 indo 10cm") in the Norwegian Deeps. Diversity N 1 (expressed in equivalent number of species) calculated on the total sample showed a similar ofthe North Sea 31 trend with latitude (Fig. 7b). This trend is most distinctive in the Southern Bight with an average of 38 species found in the southern stations off the Belgian coast and only I3 species south of the Dogger Bank. In the Northern North Sea diversity showed a tendency opposite to the density trend. The low number of species recorded between 57° and 58°N approximately coincides with the 100 m isobath. The mean individual weight (AFDW), obtained by dividing total biomass by total density showed a completely opposite trend (Fig. 8b). Towards the north individual size increased considerably, due to the gradual replacement of interstitial by large epibenthic species. The mean ash-free dry weight of the nordic species was nearly three times the value for the Southern Bight species. This difference in AFDW combined with the latitudinal trend displayed by density explains why total biomass reaches a peak in both the south and the north (Fig. 8a). In the Southern Bight low individual size and weight is compensated by maximum values for density; in the northern North Sea large, epipelic (= mud-dwelling) species with strongly chitinized exoskeletons (e.g. Cletodidae) occurred in low to very low numbers. Classification A TWINSPAN run with standard options resulted in a dendrogram with 34 clusters of which four were onesample clusters (Fig. 9). A total of 877 species and pseudospecies were recognized. The first dichotomy separated the deepwater samples from the shallow stations. Since depth is only partially correlated with latitude, the separation between the two station-groups coincided with a southwest-northeast boundary rather than with an east-west isobath. The first cluster groups 110 stations located in the entrance to the Firth of Forth (Scotland) and in the Southern North Sea, i.e. south and east of the Dogger Bank, including the Dogger Bank stations. Indicator species for this cluster were two small-sized Leptastacidae Leptastacus laticaudatus and Paraleptastacus espinulatus. Fifty-five stations were grouped in the second cluster. The most highly preferential species for this deepwater cluster were the stenocopiid Anoplosoma sordidum and the continental shelf cerviniid Cerviniopsis clavicornis. Seven main twin groups, based on ordination, can be derived from Figure 9. The second dichotomy of the shallow water cluster divided the stations in two secondary clusters of equal size (55 stations) corresponding to twin groups a, TWIN A and TWIN B, and to TWIN C and B, respectively. The secondary division of the deepwater cluster results in two more twin groups TWIN D and TWIN E. No genuine indicator species could be identified for the cluster combining TWIN A, TWIN B and the one-station cluster a. Cluster a constituted the first offshoot in the latter cluster and represented station 37 only (Fig. 9). This locality is situated in the Dutch Wadden Sea (Fig. 10) and apparently is Table 2. Individual ash-free dry weight (llg) and number of specimens weighed (n) for 98 copepod species. AFDWind.-' (ug) Arneiridae Arneirinae Ameira longipes Ameira parvula Ameira tenella Ameiropsis brevicornis Ameiropsis mixta Interleptomesochra eulittoralis Proameira hiddensoensis Pseudameira erassicornis Pseudameira mixta Sarsameira exilis Sarsameira parva Sicameira leptoderma Stenocopiinae Anoplosoma sordidum Malacopsyllus fragilis Stenocopia longicaudata Ancorabolidae Ancorabolus mirabilis Canuellidae Canuella perplexa 1.14 0.91 0.79 1.20 0.73 0.38 0.79 0.71 0.65 3.05 0.65 0.52 25 35 40 50 50 45 30 36 42 35 30 28 1.96 2.25 3.16 33 29 45 0.91 23 4.82 50 Cerviniidae Cervinia bradyi Cervinia synarthra Cerviniopsis clavicornis Eucanuella spinifera 7.33 6.05 7.70 6.21 20 20 20 20 Cletodidae Argestes mollis Cletodes limicola Enhydrosoma garienis Rhizothrix curvata 6.40 1.21 0.91 0.79 25 35 45 38 Cyclopinidae Metacyclopina 0.28 50 Cylindropsyllidae Boreopontia heipi Cylindropsyllus laevis Cylindropsyllus remanei Evansula pygmaea Leptopontia curvicauda Stenocaris minor 0.41 1.29 0.58 0.51 0.39 0.61 55 31 25 65 52 40 Diosaccidae Amphiascus minutus Bulbamphiascus imus H aloschizopera pygmaea Paramphiascoides vararensis Paramphiascopsis longirostris Psammotopa phyllosetosa Pseudamphiascopsis herdmani Pseudomesochra longifurcata Stenhelia gibba Typhlamphiascus con/usus 1.06 4.48 0.93 2.25 3.06 0.47 7.02 0.94 0.87 4.25 35 22 36 25 30 55 18 28 25 35 Ectinosomatidae Arenosetella germanica Bradya typica Ectinosoma melaniceps Halectinosoma gothiceps Halectinosoma herdmani Halectinosoma propinquum Halectinosoma sarsi Pseudobradya beduina Pseudobradya minor 0.35 4.27 1.48 1.11 1.53 2.98 3.15 1.45 1.45 65 27 35 45 40 26 35 48 50 brevisetosa AFDWind.-1 (ug) n n Laophontidae Asellopsis hispida Asellopsis intermedia Heterolaophonte stroemi Laophonte corn uta Laophonte elongata Laophonte inopinata Laophonte thoracica Paralaophonte congenera Pseudolaophonte spinosa Pseudonychocamptus proximus 1.65 1.61 3.60 3.75 2.65 3.06 2.45 2.45 5.87 3.81 Leptastacidae Arenocaris bijida Leptastacus laticaudatus Leptastacus sp. 6 Paraleptastacus espinulatus Paraleptastacus holsaticus 0.26 0.27 0.78 0.22 0.41 75 65 35 100 100 Longipediidae Longipedia coronata Longipedia helgolandica Longipedia minor 6.87 5.89 5.71 25 25 25 N ormanellidae Normanella rnucronata 1.51 38 0.23 0.18 0.25 0.18 0.39 0.17 0.56 0.19 0.14 0.18 0.20 0.24 0.28 0.31 0.22 0.58 85 50 65 70 46 85 74 125 150 100 72 114 133 78 108 65 Paranannopidae Danielssenia typica Psammis longisetosa 2.28 1.96 35 26 Tachidiidae Euterpina acutifrons Microarthridion littorale 1.83 1.46 20 65 Tetragonicipitidae Pteropsyllus consimilis 3.10 15 Thalestridae Dactylopusia vulgaris Pseudotachidius coronatus 2.00 4.33 65 25 Thompsonulidae Thompsonula hyaenae 1.96 34 Tisbidae Tisbe furcata 1.20 25 Zosimidae Tachidiella minuta Zosime major Zosime typica 1.11 1.39 1.21 29 48 55 Paramesochridae Apodopsyllus listensis Apodopsyllus spinipes Apodopsyllus sp. I Diarthrodella secunda Gen. I sp. I Gen. 2sp. I Intermedopsyllus intermedius Kliopsyllus constrictus Kliopsyllus holsaticus Kliopsyllus paraholsaticus Leptopsyllus elongatus Paramesochra helgolandica Paramesochra mielkei Paramesochra similis Scottopsyllus minor Wellsopsyllus gigas 35 28 35 45 34 37 26 31 20 26 Meiobenthos 250 of the North Sea 130 120 (a) 200 33 (a) 110 100 '" 'E 150 90 u ~ 80 100 70 60 50 50 40 o '1' lJ 51 30 51 59 Q 40 Cl' (b) ::l.. 35 3.5 ( b) 30 3 25 2.5 20 15 2 10 1.5 5 o 51 52 53 54 55 58 59 Latitude Figure 7. Copepoda. Trend of density (a) and diversity N 1 (b), calculated on the total sample, with latitude. Per degree latitude density and diversity of all stations falling in that zone is averaged. Error bars indicate standard errors of the mean. influenced by the low salinity in this area. The only (indicator) species found is Tachidius discipes, a typical representative of estuarine and near-coastal localities where lower salinities prevail. Due to their distribution pattern (Fig. l Id) certain Leptastacidae such as Arenocaris bijida and Leptastacus laticaudatus can be considered as preferential (together with the fusiform burrower Halectinosoma herdmani) for the joint cluster [TWIN A-TWIN B]. Another leptastacid Paraleptastacus espinulatus is an indicator species for the group. TWIN A. TWIN A represents perhaps the most distinctive community found in the North Sea. It consists of 22 highly diverse stations located in the area traditionally referred to as the Southern Bight (except stations 92, 155 and 166) and roughly demarcated in the north at 53.SON (Fig. 10). This area is bounded by the Belgian coast and Dutch Delta area in the east and the coast of Norfolk in the west, but excludes The Wash and the coastal area off the rivers Thames and Rhine/Meuse, Overall, the Southern Bight consists of fine « 250 urn) to medium coarse (250--500 urn) sandy sediments with a low silt clay content reaching a maximum of2.6% in station 19. Copepod densities ranged from 24 to 651 (x = 178) indo 10 em -2, which are the highest observed in this study. On the 0.5 51 52 53 54 55 56 57 58 59 Latitude Figure 8. Copepoda. Trend of total biomass (a) and mean individual ash free dry weight (b), with latitude. Per degree latitude the biomass and individual AFDW of all stations falling in that zone is averaged. Error bars indicate standard errors of the mean. average copepods accounted for 25% of the meiobenthos, occasionally contributing as much as 50% of the total density at some stations. The Southern Bight (19 stations) contains approximately 50% of the total number of species found during the North Sea Benthos Survey. Indicator species for this twin group are Kliopsyllus holsaticus, Leptopontia curvicauda, Intermedopsyllus intermedius, Evansula pygmaea and Me tacyclopina brevisetosa, Nearly all other representatives of the families Paramesochridae, Cylindropsyllidae and Cyclopinidae were selected as preferential species. The great majority of the fauna was made up by mesopsammic ( = interstitial) species inhabiting the space between the sand particles. These animals crawl or swim within the lacunae with no, or only negligible, disturbance to the structure of the sediment. The harpacticoid species from this area can be said to have adapted to the interstitial habitat primarily by miniaturization of the body (Paramesochridae, Cyclopinidae) or adoption of extreme vermiformicity (cylindrical shape) and reduction of the appendages (Cylindropsyllidae). The Southern Bight assemblage posed great difficulties as to accurate identification because nearly half of the species (61) were new to science. It contained all North Sea Cylindropsyllidae (Fig. llc) and Leptastacidae (Fig. lId), and all but one R. Huyset 34 TWIN A TWIN B • * TWIN ~- TWIN TWIN C o E X J. • al. A similar community dominated by interstitial species was found in two localities along the Danish coast (Fig. 10). Station 92 (427 indo 10em -2) is situated near the Isle of Sylt. Mielke (1975) already reported on a diverse mesopsammic assemblage from the coarse sandy sediments in this area whilst Herbst (1974) described several interstitial cyclopinids from the same deposits. The second station (155) is associated with the Jutland Bank in the north; the coarse sandy sediment is inhabited by a high number (390 indo 10em -2) of interstitial species. However, Paramesochridae were less abundant and were primarily replaced by vermiform Ectinosomatidae and Diosaccidae. Finally, TWIN A also included a third station (166) that is geographically isolated from the Southern Bight. This Firth of Forth station was separated first from the rest of TWIN A because of the mixed species composition made up of Leptastacidae and epibenthic faunal elements (Laophontidae); the absence of other interstitial species was remarkable. TWIN B. TWIN B could not be identified by any indicator species nor did the analysis point to any preferential a ~ species. This twin group essentially coincided with: (i) the [0 15, 8, II, 22[ zone along the eastern coastline of the Central North Sea, Figure9. TWINSPANdendrogrambasedon the speciescompo- extending from the Terschelling Bank in the south, over sitionof the Copepoda(* indicatesl-sampleclusters). the German Bight and the Danish westcoast, to the entrance to the Skagerak in the north, (ii) the localities (38,39,40,50) north of The Wash between 53° and 54°N, (Wellsopsyllus gigas) of the 44 recorded Paramesochridae species (Fig. Ila). Representatives of these three families and (iii) the shallow stations located at the Dogger Bank (Fig. 10). The latter two regions are separated by a comwere associated with small Ameiridae Unterleptomesplex of deep trenches ("Silver Pits") which harbour a difochra, Leptomesochra, Parevansula, Sicameira) and verferent copepod fauna (see TWIN D). In general, shallow miform Diosaccidae (Psammotopa, Protopsammotopa) and Ectinosomatidae (Hastigerella, Arenosetella). A typi- stations with fine to very fine sandy sediments with a cal element of the Southern Bight community was the low amount of silt and clay were grouped in this cluster. cyclopoid family Cyclopinidae (Fig. II b). Due to their The fauna was characterized by a mixture of minute adaptations to the mesopsammic lifestyle, these tiny interstitial species and large burrowing forms. The cyclopoids (Metacyclopina, Cyclopina, three new mesopsammic component is dominated by Leptastacidae (Leptastacus, Paraleptastacus, Arenocaris) which is the genera) show a remarkable convergence with the paramesochrid harpacticoids. In many stations the inter- only exclusively interstitial family that extends into this zone (Figs lid, 12). Conversely, Cylindropsyllidae, stitial community was accompanied by characteristic Cyclopinidae and Paramesochridae were completely coarse sediment-inhabiting species such as Rhizothrix absent north of TWIN A (Fig. 12); they were replaced by spp. (Cletodidae) and various Tetragonicipitidae a few interstitial representatives of the families Ectino(Pteropsyllus, Tetragoniceps). The species composition strongly resembled the somatidae and Ameiridae whose relative abundances mesopsammic assemblage of the coarse sands of the were low. The majority of the community, however, Kwinte Bank as described by Willems et al. (1982). The consisted of fusiform Ectinosomatidae (Ectinosoma, similarity between the harpacticoid associations from the Halectinosoma, Pseudobradya) and, to a lesser extent, Ameiridae (Ameira, Proameira, Pseudameira). Densities Southern Bight and that of the coarse sands ofthe French Catalonian coast (Soyer, 1970) and the coarse sand were low, ranging from 3-81 indo 10em -2. association of the Irish Sea (Moore, 1979) suggests that the copepod faunas of medium and coarse (> 300 11m) TWIN C. TWIN C reflects a transition community offshore deposits are similar, provided that the sands are between the coastal Ectinosomatidae-Leptastacidae well-sorted and clean. Comparison with earlier studies in association (TWIN B) and the deepwater community the Southern Bight also reveals that the community is (TWIN D) north of the Dogger Bank. A few stations are located east of the Firth of Forth where they are interstable in time (e.g. Huys et al., 1986). Meiobenthos of the North Sea (A) (8 ) w_ Figure 10. Distribution of the five main TWINSPAN station-groups in the North Sea. For each twin group the most important ecotypes are illustrated. spersed with stations belonging to the second twin group (Fig. 10). This station-group consisted of 51 stations and was impoverished both qualitatively and quantitatively. A clearly preferential species for this twin group was Paramphiascopsis longirostris, whilst interstitial copepods were completely absent. Total densities ranged between 5 and 45 indo 10cm", The fauna consisted oflarge pelophilie ( = mud-dwelling) species belonging to the Diosaccidae (Paramphiascopsis, Stenhelia, Bulbamphiascus, ... ), Laophontidae (various genera) and Ameiridae (Ameiropsis, Pseudameira, Sarsameiray. Within the Ectinosomatidae, the larger species were still important with the smaller 35 R. Huyset 36 ( a ) al. (b ) " o-ITJ-o 174 1-1_ ----0 4 21 • o o---CO- 0 203 151 • • Figure 11. Log of total density of the major interstitial families of the Copepoda in the North Sea. (a) Paramesochridae (filled circles denote records of Wellsopsyllus gigas). (b) Cyclopinidae. (c) Cylindropsyllidae. (d) Leptastacidae. Between the minimum and maximum values indicated under the box and whisker plot, the radius of the symbols is proportional to the log-transformed density (see text for details). burrowing forms being replaced by bigger representatives of the same genera (Halectinosoma, Pseudobradya) or of Bradya. The polyarthran genus Longipedia was abundant throughout the transition zone (Fig. l3c) with L. minor and L. helgolandica widely distributed in the south and L. coronata and L. scotti in the north. Meiobenthos 90 c: 80 70 0. 60 ·2 E 37 paramesochrid Wellsopsyllus gigas (Fig. Ila), originally described from the Fladen Ground (Wells, 1965), and the diosaccid Typhlamphiascus confusus. The deep Silver Pits (stations 49, 56, 57) formed an isolated subcommunity with low densities and an impoverished fauna consisting exclusively of Zosimidae and Idyanthidae. 100 .2 of the North Sea 8 50 g;, .E 40 c: ~ 30 (L 20 :;; 10 o ABC 0 E Figure 12. Composition of the North Sea Copepoda per twin group showing importance of exclusively interstitial families in southern twin groups. ~ Cylindropsyllidae, • Cyclopinidae, i§:lLeptastacidae, § Paramesochridae, 0 others. The faunistic picture of the interstitial Southern Bight community is disturbed by the river plumes of the Wash, the Humber and the Thames (Fig. 10). Stations (5, 8, 11, 22) located in the respective river outlets were devoid of interstitial species and cluster together in group ~ (Fig. 9). Instead, low densities and few species were recorded. Station 13, influenced by the Rhine-Meuse estuary, apparently belonged to the same cluster, however, it was omitted in the TWINSPAN analysis because no copepods were encountered. Microarthridion littorale and/or Canuella perplexa were (strict) preferential species in all the stations. Earlier studies revealed a similar situation for the Westerschelde estuary (e.g. Huys et al., 1986). TWIN D. TWIN D is a heterogeneous cluster of 48 stations and coincides with the northern part of the North Sea, situated between the Norwegian Deeps and the Scottish coastline (Fig. 10). No proper indicator species could be identified for the group, but the most important families, both in terms of diversity and density, were the Cletodidae, Zosimidae and Idyanthidae. These three families nearly always occurred together in every station of the area. The Cletodidae were represented by various species of the genera Cletodes, Enhydrosoma and Stylicletodes which are typical faunal elements of deep (40-80 m), muddy bottoms. The Zosimidae iZosime, Tachidiellai occupy the same depth range (Fig. 13a) and like the Cletodidae are adapted for an endopelic existence, i.e. shallow burrowing in muddy substrates. The Idyanthidae (Fig. 13b), on the other hand, is characteristic of the flocculent upper layer and encompasses epibenthic genera (Idyella, Idyanthe, Tachidiopsisy, This assemblage was also found in stations 99 and 120 which were located in a deep trench penetrating the transition zone (TWIN C) and coinciding with the incision of the Pleistocene River Elbe estuary. Finally, two characteristic species for this area were the giant mud-dwelling TWIN E. TWIN E corresponds to the northeastern part of the study area (Norwegian Deeps). The deepwater stations 184, 185, 186, 195 and 196 can be allocated to this region where the depth varies between 84 and 100 m. To a large extent the sediments consisted of fine to medium sand. The silt clay content of the sediment over most of this area ranged between 1.3 and 3.3%, reaching a maximum of 12.4% in the deepest station 196. Copepods occurred in densities ranging from 23-128 indo 10 em -2. The community was dominated by the families Cletodidae and Ancorabolidae, and by the deepwater species Pseudotachidius coronatus (Pseudotachidiinae). The genera Eurycletodes, Mesocletodes, Argestes and Heteropsyllus accounted for more than 90% of the Cletodidae and seemed to have replaced the genera Cletodes, Stylicletodes and Enhydrosoma of the adjacent area (TWIN D). The Ancorabolidae, consisting exclusively of Ancorabolinae, is a typical component of the deepwater fauna of fjords and was represented here by three species (Ancorabolus mirabilis, Echinopsyllus normani and Ceratonotus pectinatus); they represented > 35% of the copepod fauna in most of the stations (Fig. 14b). Other characteristic faunal elements of this assemblage, not found in any of the others, were the deepwater Stenocopiinae (Ameiridae) (Fig. 14a) represented by three genera tStenocopia, Malacopsyllus, Anoplosoma), and the Cerviniidae (Fig. 14c) represented by the continental shelf genera Cervinia, Eucanuella and Cerviniopsis. In contrast to the previous community, the relative abundance of Zosimidae and Idyanthidae was negligible. Typhlamphiascus gracilis seemed to have replaced T. confusus in this deeper area; this phenomenon was also recorded by Por (I 964a). It is worthy of note that an analogous community was found in other stations widely separated geographically from the Norwegian Trench (Fig. 10). Both stations had a sediment consisting of very fine sand with a silt clay content being in excess of9%. Station 137 (91 m) is located in the Devil's Hole, a deep extension of the Fladen Ground, penetrating the Central North Sea. Its copepod fauna was a mixture of "Nordic" cletodid genera, cerviniids and Stenocopiinae. Station 103 was the deepest locality sampled during the North Sea Benthos Survey (107 m) and is situated in the Farne Deep, a depression off the coast of Northumberland. A similar fauna was found here; however, the Stenocopiinae were replaced by Ancora bolinae. R. Huyset al. 38 (a ) ( b) O-DJ-O 3 5 • •• O-ITJ-I 3[ 20 • • • • •• • (c ) 3 o-[]_ro [0 • Figure 13. Log of total density of some important copepod families in the Central North Sea. (a) Zosimidae. (b) Idyanthidae. (c) Longipediidae. Between the minimum and maximum values indicated under the box and whisker plot, the radius of the symbols is proportional to the log-transformed density (see text for details). Meiobenthos (a) 39 of the North Sea ( b ) 4 6 0[/ ~o 27 3 0-CO--0 22 3 • • ( c) 3 19 • Figure 14. Log of total density of typical deepwater copepod families in the North Sea. (a) Ameiridae (Stenocopiinae). (b) Ancorabolidae. (c) Cerviniidae. Between the minimum and maximum values indicated under the box and whisker plot, the radius of the symbols is proportional to the log-transformed density (see text for details). R. Huysetal. 40 Table 3. Mean density (n. 10em -2), number of species, diversity N 1> biomass (ug, 10em "') and individual ash-free dry weight (ug, indo-1) of Copepoda per twin group. The averaged total copepod density per twin group is illustrated in Figure l5a. TWIN A is clearly separated by density which was about 4 to 10 times the mean density of the other groups (Table 3). The central twin groups TWIN B-D are very impoverished. 250 (a) 200 Twin A TwinB TwinC TwinD TwinE Density No. of species exp (H') 192 29.1 30 10.9 7.8 8.9 9.7 22.8 8.9 6.6 7.5 85.7 28.6 7.1 158.9 25 19 51 Biomass 66.1 32.1 Individual AFDW 0.52 0.77 3.00 1.57 3.17 N 'E u o a z 250 (a) 200 o 35 N 'E u 30 Q 0> Ci E g ~ 0> ::L 100 25 0> Cl. Of) 0> 150 50 20 'u 0> o, Of) 's a z 15 10 3 5 2.5 30 (c) 25 2 1.5 Of) 0> 'u 0> Cl. 20 Of) ~o ac 0.5 15 o 0w 10 A B c D E Figure 16.Averaged total biomass (a) and individual ash-free dry weight (b) of Copepoda per twin group. 5 Figure 15. Averaged total density (a), number of species (b) and diversity N (c) of Copepoda per twin group. j Species diversity per 10 ern - 2 followed the trend of density (Figs l5b, c). The number of species in TWIN A was at least three times higher than in the other twin groups. The lowest mean diversity was recorded in transition group TWIN C. Species abundances per twin group have been converted to biomass values (Fig. l6a) using the data of Table 2. Total biomass was highest in the deepwater twin group E, followed by TWIN A and TWIN C. Individual dry weight was also highest in TWIN E, but the difference with other groups was not as pronounced as for total biomass (Fig. l6b). The copepods of the three northern twin groups had the highest mean individual size, those of TWIN A by far the smallest, TWIN C and D had comparable individual AFDW values. Ordination Canonical correspondence analysis was performed on the five major twin groups A-E. Stations belonging to Meiobenthos of the North Sea clusters a and p were omitted in the analysis because granulometric data were not available. Environmental variables used are depth, latitude, longitude, percentage silt, percentage clay and median grain size of the sand fraction. Silt percentage, clay content and median grain size (in <D-units)are highly correlated (Fig. 17) and there seemed to be almost no correlation between these variables and depth. Latitude was more closely related to depth than to sediment characteristics; there is a general trend of increasing water depth to the north. 41 the Irish Sea and bounded in the north and the west by the 200 m depth contour, in the south and east by the continental coastline from Brest to Bergen) is estimated at 850. The difference in species number can partly be explained by not having sampled the phytal environment which is known to harbour a species-rich community. It foreshadows an explosive increase in number of species as sampling effort increases. Bearing in mind the unexpected diversity in coarse sediments and the virtual lack of knowledge from the northern North Sea, the total number of benthic copepods harboured by North Sea sediments might safely be estimated at at least 1500 species. 400 300 Nematode/copepod ratio ··········r··········f ..········ 200 o -100 -200 ! , _._._~--- . . . -300 ·········,··········r····· ..··· ···········i··········t -400 -300 -200 -100 0 ..... , ! tp. [ ~ 100 200 300 400 500 600 . 700 Figure 17. CCA ordination diagram of the five major twin groups based on copepod composition. The identity of the five major twin groups is clearly expressed in the ordination diagram, with no overlapping between the respective station groups. The three southern twin groups A, Band C show a close correlation with the sediment and can be arranged along a gradient of decreasing grain size. The coarsest sediments are grouped in TWIN A, the finest in TWIN C. The position in the biplot of the northern station groups D and E is obviously related to latitude and to depth. Stations belonging to TWIN E are grouped together mainly because of their great depth and their high latitude; they are clearly separated from the TWIN D station group. Discussion The fauna The percentage of new species recorded during the North Sea Benthos Survey is overwhelming and unexpected. Nearly 44% ofthe 278 species are new to science. A survey of the ecological literature on Harpacticoida of the Belgian-Dutch continental shelf produced a total number of276 species. The number of species recorded in the area delimited for the Synopses of the British Fauna (including Since Raffaelli & Mason (1981) made the attractive suggestion of using the ratio of freeliving nematodes to benthic copepods as a practical pollution indicator for sandy beaches, the literature on using meiobenthos in pollution monitoring has been fueled with controversy (see Lambshead, 1984 for review). Most authors have now abandoned the NjC ratio because it oversimplifies the complex responses ofmeiobenthos populations to the environment. The North Sea Benthos Survey provided the opportunity to assess the potential of the NjC ratio as a pollution indicator on a large geographical scale (Fig.4a). It is conceivable that meiobenthic populations are most highly subjected to anthropogenic pollution in the Southern Bight. Yet, the NjC ratio is remarkably low in this area, even if only interstitial copepods are included in the estimation of the NjC ratio, and the copepod distribution does not suggest that the southern North Sea is more polluted than the northern North Sea. The fact that there is almost an order of magnitude of disparity between the Southern Bight and the rest of the North Sea (Fig. 4b) makes it unnecessary to invoke pollution as an explanation. On the other hand the NjC ratio varies considerably in the central and northern North Sea, suggesting that nematodes and copepods are influenced independently by a complex suite of environmental parameters. Most likely, the different habitat requirements of nematodes and the two major groups of copepods in relation to the granulometry of the sediment account for this variability. Hence, it is impossible to apply adequately the NjC ratio once the study area is extended to habitats other than sandy beaches (Raffaelli, 1987). Communities The topography of the North Sea is an important factor in determining the pattern of water movements and thus the environmental conditions to which the animals and plants in the various areas are subjected. Adams (1987) recognized seven subdivisions of the North Sea by using certain R. Huysetal. 42 30 2° 1° 0° 30 2° 1° 0° 1° 2° 3° 4° 5° 6° 7° 8° 9° 1° 2° 3° 4° 5° 6° 7° 8° 9° (b) (a) 61° 60° 59° I .' ~: 61° 60° 59° 58° 57" 56° 54° 53° Figure 18. Comparison between Adams' (1987) primary subdivisions based on copepod composition (b). of the North Sea (a) and major twin groups (indicated by A to E) depth contours and a combination of physical and bio- origin, is delimited by the 100m and 50 m isobaths in the logical properties of the water masses (Fig. 18a). Adams north and south, respectively, and can almost be identused the 40 m contour, which appoximately marks the ified by TWIN D. In the west the depth varies from 50 to boundary between stratified and well mixed water during 80 m. Finally, the "Norwegian Deeps" coincide with the summer, the 50 m contour, which along part of its TWIN E. The "Offshore Northern Zone" was not course coincides with northern flank of the Dogger Bank, sampled during the North Sea Benthos Survey. In contrast to nematodes (e.g. Vincx, 1989;Vincx et al., the 100 m contour, along which the water masses of the Fair Isle-Orkney current tend to flow, and the 200 m 1990; Vanreusel, 1990), harpacticoids have been less contour, which marks the western boundary of the frequently used to describe North Sea communities, the only comprehensive study being that by van Damme & Norwegian Deeps. There is a certain resemblance between Adams' (1987) Heip (1977). Three distinct zones (slightly modified by subdivisions and the twin groups obtained in this study Govaere et a!., 1980)were classified in the Southern Bight (Figs 18a, b). The shallow water cluster made up by according to their harpacticoid copepod composition: the coastal zone with a Microarthridion littoraleTWIN A and B largely coincides with the "Continental Coastal Zone" whose offshore boundary follows the 40 m Halectinosoma herdmani community defined by large epidepth contour except in the southwest where it marks the benthic and endobenthic species, the transition zone chareastern extent of the "South British Coastal Zone". The acterized by a Leptastacus laticaudatus-Halectinosoma latter area was rather arbitrarily defined by Adams (1987) herdmani community, and the open sea characterized by helgolandica and was not recognized as such in the present study. It the Leptastacus laticaudatus-Paramesochra might nevertheless be related to cluster ~ which groups the community. The interstitial assemblage of the "open sea" river plume stations. The "Offshore Southern Zone" coin- as defined by these authors clearly coincides with the cides with TWIN C except for the Dogger Bank proper community described for TWIN A. The stability in time and space of this Southern Bight community is at least and the Silver Pits which belong to TWIN B and TWIN D, respectively. This region lies between the 40 m and remarkable. Govaere et a!.'s (1980) conclusions were 50 m depth contours and its water column becomes strat- based on samples collected from 1970 until 1975. Over a time span of two decades the only significant difference ified during the summer only. The "Offshore Central Zone" whose water mass is mainly North Atlantic in found is the dominance ranking of P. helgolandica; this Meiobenthos of the North Sea 43 It is conceivable that similar observations in the future species seems now to be replaced by various other Paramesochridae (Table I). The coastal and transition zones will contribute to our understanding of the distribution patterns of the various copepod families. Sediment were not recognized in the present survey but that might characteristics, depth and latitude are obvious variables merely be a result from the choice of the sampling grid. The transition zone community can be identified with to account for the structure ofmeiobenthic communities, however, the underlying biological reasons, which are of TWIN B (mixture of Leptastacidae and large burrowers) paramount importance, are still not well understood. but did not extend as far south as shown by Govaere et al. (1980: Fig. 3). The description ofmeiobenthos distribution in terms of Acknowledgements communities is a well-established exercise and dates back to the thirties when Remane (1933) used meiobenthic The authors thank the following scientists who particispecies to define benthic communities in the Kiel Bay. pated in the North Sea Benthos Survey: U. Niermann Many of the subsequent studies (e.g. Por, I 964b; Coull & (Biologische Anstalt, Helgoland), T. Brey and H. Herman, 1970; Soyer, 1970) attempted to apply the Rumohr (Universitat Kiel), A. Kiinitzer and E. Rachor isocommunity concept introduced by Thorson (1957) for (Alfred Wegener Institut, Bremerhaven), J. Dorjes macro benthic communities. This concept was based on (Senckenberg Institut), G. Duineveld and P. de Wilde the assumption that communities inhabiting the same (NIOZ, Texel), P. Kingston (Heriott-Watt University, type of bottom at similar depths are characterized by difEdinburgh), and J. M. Dewarumez (Institut de Biologie ferent species of the same genera, but replacing each other Marine, Wimereux). We also would like to acknowledge in accordance with the geographical regions. This hypoththe invaluable assistance of the crews on the vessels proesis has not stood the test of time and the present survey vided by the Universitat Kiel, the Biologische Anstalt clearly showed that depth and sediment type are not the Helgoland, the Senckenberg Institut, the Institut fur only factors that structure harpacticoid communities. For Meeresforschung Bremerhaven, the NIOZ, the Dutch example, the Southern Bight community inhabiting clean Rijkswaterstaat and the Belgian Ministry of Public sandy bottoms is not found in similar deposits at similar Health. Grateful thanks are due to M. P. W. J. Braat for depths in the entrance to the Skagerak. This might indistatistical assistance. The senior author acknowledges a cate that either a major physical variable was not CEC Science grant no. ST2*0443. measured, or that the species assemblage was affected by biological interactions that were not investigated. For References example, grain size has often been regarded as the most significant parameter influencing the distribution ofmeioAdams, J. A. 1987. The primary ecological subdivisions of the benthos. This led Wieser (1959) to introduce a minimum North Sea: some aspects of their plankton communities. In critical grain size of approximately 200/lm as a barrier Development in Fisheries Research in Scotland. Ed. by R. S. Bailey and B. B. Parish. London, Gishing News, pp. 165-181. between interstitial sliders and burrowing species. Pennak Anonymous, 1986. Fifth Report of the Benthos Ecology (1950) on the other hand stated that the size of sand grains Working Group. International Council for the Exploration of has no constant relationship to either distribution or the Sea 1986/C:27. number of organisms. The distribution of the interstitial Basford, D. J., Eleftheriou, A., and Raffaelli, D. 1989. The epifauna of the Northern North Sea (56°-61 ON).Journal of the families of Copepoda in the North Sea (Fig. 11) gives Marine Biological Association of the UK, 69:387-407. evidence that at least for the Leptastacidae grain size Basford, D. J., Eleftheriou, A., and Raffaelli, D. 1990. The per se does not affect their distribution. The genera infauna and epifauna of the northern North Sea. Netherlands Leptastacus and Paraleptastacus which are generally Journal of Sea Research, 25:165-173. regarded as interstitial sliders (Wells, 1986) do not seem to Coull, B. c., and Herman, S. S. 1970. Zoogeography and parallel level-bottom communities of the meiobenthic Harpacticoida be affected by the amount of silt which might fill up the (Crustacea, Copepoda) of Bermuda. Oecologia (Berl.), 5: interstitial spaces of coarser sands, nor by the actual grain 392-399. size of the sand fraction. The Leptastacidae are the only Eleftheriou, A., and Basford, D. J. 1989. The macrobenthic vermiform copepods whose distribution extends to the infauna of the offshore Northern North Sea. Journal of the Marine Biological Association of the UK, 69: 123 -143. Central North Sea. One of the possible factors explaining Gee, J. M., and Warwick, R. M. 1984. Preliminary observations this distribution pattern might be sought in their feeding on the metabolic and reproductive strategies of harpacticoid biology. Live observations showed that their feeding copepods from an intertidal sandflat. Hydrobiologia, 118: strategy is rather different from typical grazing as dis29-37. played by many interstitial copepods (e.g. CylindropsylGoodman, K. S. 1980. The estimation of individual dry weight and standing crop of harpacticoid copepods. Hydrobiologia, lidae, Paramesochridae) and does not necessarily depend 72:253-259. on sediment type since food particles are passively trapped Govaere, J. C. R., van Damme, D., Heip, c. and De Coninck, in the mucus strands produced by the caudal glands. It also L. A. P. 1988. Benthic communities in the Southern Bight explains why Leptastacidae were sometimes found to live of the North Sea and their use in ecological monitoring. Helgolander Meeresuntersuchungen, 33:507-521. in co-existence with genuinely pelophilic harpacticoids. 44 R. Huyset al. Heip, C. H. R .., Huys, R., Vincx, M., Vanreusel, A., Smol, N., Herman, R., and Herman, P. M. J. 1990. Composition, distribution, biomass and production of North Sea meiobenthos. Netherlands Journal of Sea Research, 26:333-342. Heip, c., Vincx, M., and Vranken, G. 1985. The ecology of marine nematodes. Oceanography and Marine Biology Annual Review, 23:399-489. Herbst, H.-V. 1974. Drei interstitielle Cyclopinae (Copepoda) von der Nordseeinsel Sylt. Mikrofauna Meeresboden, 35: 1-17. Herman, P. M. J., and Braat, M. P. W. J. 1991. DIHOMAPManual. Internal report, Delta Institute for Hydrobiological Research. Herman, P. M. J., and Heip, C. 1985. Secondary production of the harpacticoid copepod Paronychocamptus nanus in a brackish water habitat. Limnology and Oceanography, 30: 1060-1066. Herman, P. M. J., Heip, c., and Guillemijn, B. 1984. Production of Tachidius discipes (Copepoda: Harpacticoida). Marine Ecology Progress Series, 17:271-278. Hill, M. O. 1973.Diversity and evenness: a unifying notation and its consequences. Ecology, 54:427-432. Hill, M. O. 1979. TWINSPAN - A Fortran program for arranging multivariate data in an ordered two-way table by classification of the individuals and their attributes. Ecology and Systematics. Cornell University, Ithaca, New York, 48 pp. Huys, R. in press. The amphiatlantic distribution of Leptastacus macronyx (T. Scott, 1892) (Copepoda: Harpacticoida): a paradigm oftaxonomic confusion; and, a cladistic approach to the classification of the Leptastacidae Huys. Academiae Analecta, in press. Huys, R., Vanreusel, A., and Heip, C. 1986.The meiobenthos of the Voordelta. Final Report project BOVO, Rijkswaterstaat, Nederland, 88 pp. [in Dutch]. Lambshead, P. 1. 1984. The nematode/copepod ratio. Some anomalous results from the Firth of Clyde. Marine Pollution Bulletin, 15:256-259. McIntyre, A. D. 1964.Meiobenthos of sublittoral muds. Journal of the Marine Biological Association of the UK, 44:665-674. Mielke, W. 1975. Systematik der Copepoda eines Sandstrandes der Nordseeinsel Sylt. Mikrofauna Meeresboden, 52:1-134. Moore, C. G. 1979.Analysis of the associations ofmeiobenthic Copepoda of the Irish Sea. Journal of the Marine Biological Association of the UK, 59:831-849. Pennak, R. W. 1950. Comparative ecology of the interstitial fauna of fresh-water and marine beaches. Colloque Internationaux du Centre National de Ia Recherche Scientifique, Paris, 1950:449-480. Por, F. D. 1964a. Les Harpacticoides (Crustacea, Copepoda) des fonds meubles du Skagerak. Cahiers de Biologie Marine, 5: 233-270. Por, F. D. I964b. A study of the Levantine and Pontic Harpacticoida (Crustacea, Copepoda). Zoologische Verhandelingen (Leiden), 64: 1-128. View publication stats Raffaelli, D. 1987.The behaviour ofthe nernatode/copepod ratio in organic pollution studies. Marine Environmental Research, 23:135-153. Raffaelli, D. G., and Mason, C. F. 1981. Pollution monitoring with meiobenthos, using the ratio of nematodes to copepods. Marine Pollution Bulletin, 12:158-163. Remane, A. 1933. Verteilung und Organisation der benthonischen Mikrofauna der Kieler Bucht. Wissenschaftliche Meeresuntersuchungen der Kommission zur wissenschaftlichen Untersuchung der Deutschen Meere, Abteilung Kiel, N.F., 21: 161-221. Smidt, E. 1. B. 1951. Animal production in the Danish Wadden Sea. Meddelelser fra Kommissionen for Danmarks Fiskeri og Havundersogelser, Serie Fiskeri, 11(6):1-151. Soyer, J. 1970. Bionomie benthique du plateau continentale de la cote catalane francaise. III. Les peuplements de Copepodes Harpacticoides (Crustacea). Vie Milieu, Ser, B, 21: 337-511. ter Braak, C. J. F. 1986. Canonical correspondence analysis: a new eigenvector technique for multivariate direct gradient analysis. Ecology, 67:1167-1179. Thorson, G. 1957. Bottom communities (sublittoral of shallow shelf). In Treatise on marine ecology and paleoecology, vol. I, Ecology. Ed. by J. W. Hedgpeth. Bulletin of the Geological Society of America, 67:461-534. van Damme, D., and Heip, C. 1977.Het meiobenthos in de zuidelijke Noordzee. In Projekt Zee. Ed. by J. Nihoul and 1. De Coninck, 7:1-133. Brussels. van Damme, D., Heip, c., and Willems, K. A. 1984. Influence of pollution on the harpacticoid copepods of two North Sea estuaries. Hydrobiologia, 112:143-160. Vanreusel, A. 1990. Ecology of the free-living marine nematodes from the Voordelta (Southern Bight of the North Sea). I. Species composition and structure of the nematode communities. Cahiers de Biologie Marine, 31:439-462. Vincx, M. 1989.Free-living marine nematodes from the Southern Bight of the North Sea. Academiae Analecta, 51: 37-70. Vincx, M., Meire, P., and Heip, C. 1990. The distribution of nematode communities in the Southern Bight of the North Sea. Cahiers de Biologie Marine, 31:107-129. Wells, J. B. 1. 1965.Copepoda (Crustacea) from the meiobenthos of some Scottish marine sub-littoral muds. Proceedings of the Royal Society of Edinburgh, 69B:I-33. Wells, J. B. J. 1986. Copepoda: Marine-interstitial Harpacticoida. In Stygofauna Mundi. Ed. by 1. Botosaneanu. E. J. Brill/Dr W. Backhuys, Leiden, pp. 356-381. Wieser, W. 1959. The effect of grain size on the distribution of small invertebrates inhabiting the beaches of Puget Sound. Limnology and Oceanography, 4: 181-194. Willems, K. A., Vincx, M., Claeys, D., Vanosmael, C., and Heip, C. 1982. Meiobenthos of a sublittoral sandbank in the Southern Bight of the North Sea. Journal of the Marine Biological Association of the UK, 62:535-548.